[In the following excerpted introduction to his translation of the Memorable Doings and Sayings, Wardle considers the composition date, content, structure, sources, and textual history of the collection.]

1. THE AUTHOR

There is very meagre material for a biography of Valerius Maximus; all of it is internal: he tells us that he witnessed the suicide of an old woman in the town of Iulis on the island of Ceos while he was journeying to Asia in the company of Sex. Pompeius.¹ Pompeius is usually identified with the consul of ad 14, who was proconsul of Asia in the mid-20s ad, but the casual way he is mentioned would make V. seem a very ungrateful client;² Pompeius may thus become a humble unknown and any date for the episode be lost. A more allusive mention of ‘our rather small wealth’ (parvulos census nostros) has suggested a man of limited financial means, but it may well be no more than a human generalization.³ Certainly, though, he addresses the emperor Tiberius with abject humility.⁴ Possible connections with the patrician Valerii have been surveyed, since V. bears their grandiose cognomen Maximus, but no more than conjecture is possible.⁵ That he followed no public career is almost certain—at least no trace of one has survived.

2. THE DATE OF FACTORUM ET DICTORUM MEMORABILIUM

That the work was composed during Tiberius' reign is clear, as the references to the emperor from the preface through to the last book show.⁶ There is no indisputable reference to events from Caligula's or later reigns and Tiberius is never spoken of as dead.⁷ Since Kempf's first edition three passages have been employed to argue publication between 27 and 31, but the case is far from watertight.⁸ The first rests on the identification of V.'s Sextus Pompeius (2. 6. 8) with the proconsul of Asia, which would fix book 2 after 24/25, but as we have seen the identification is far from certain. The second passage comes from the preface to book 6: ‘For you [Chastity] dwell on the altars sacred to Vesta, you sit on the couches of Capitoline Juno, you keep a constant watch over the summit of the Palatine, the gods of the Augustan house and the most holy marriage bed of Julia …’⁹ For Kempf this is an indirect reference to Livia, whose formal name after 14 was Julia Augusta, and must predate her death in 29, whereas Carter prefers a clumsy but straightforward reference to Tiberius' ex-wife Julia and an allusion to the fact that Tiberius had never remarried. However, in addition to whatever insensitivity we may admit to V.'s words, this last interpretation seems very unlikely, because Julia's marriage bed was not protected by chastity, as her relegation for adultery made plain. If we accept the supplement of Pighius (gentis) and read ‘marriage bed of the Julian ‹family›’, there are sufficient ostensible examples of chastity in Livia, Agrippina, and Livilla for V.'s words to seem unobjectionable. Even as late as 29 there is no need to restrict the reference to Livia. As V. can refer to living female members of the imperial family straightforwardly when he wants to (4. 3. 3: Antonia), and as the name Julia never appears without a clarifying explanation of the family relationship (4. 6. 4, 6. 7. 3), ‘Iuliae’ here is best taken as an adjective (cf. 4. 5. 3) and as part of a general reference to the family, and so some supplement be required.

The third and longest passage offers greatest room for debate. V.'s reader unexpectedly encounters a fierce polemic against an unnamed individual who had failed in his attempt to assassinate Tiberius. It forms the climactic conclusion to a chapter in book 9 on scandalous sayings and wicked actions:

But why do I pursue the former examples or linger on the latter, when I see that all wicked deeds have been surpassed by the contemplation of one parricide? So with every impulse of my mind, with all the force of my indignation I am swept along to lash out at it with a dutiful rather than emotional reaction: for who, when the reliability of friendship has been destroyed, could with sufficiently powerful words consign to the depths of execration he deserves the man who tried to bury mankind in bloody darkness? Would you, clearly more savage than the monstrosity of wild barbarism, have been able to take the reins of the Roman empire, which our Princeps and Parent holds in his saving right hand? Or, with you so in the grip of madness, would the world have remained in its accustomed state? You wanted in the insane intentions of your madness, to recreate and surpass the capture of our city by the Gauls, ‹the river Cremera› fouled with the slaughter of three hundred men from one noble clan, the day of the Allia, the crushing of the Scipios in Spain, Lake Trasimene, Cannae and the madness of the civil wars dripping with Roman blood. But the eyes of the gods were awake, the stars retained their strength, the altars, couches and temples with their present divine power were protected and nothing which ought to have maintained watch over the august head and fatherland allowed itself laxness, and above all the author and defender of our safety saw to it that his own most excellent achievements would not be buried in the ruin of the whole world. So peace stands firm, the laws are in force, the course of private and public duty is preserved unimpaired. He, who by violating the bonds of friendship tried to subvert all this, along with all his offspring, has been crushed by the might of the Roman people and is paying the punishment he deserves in Hell, if he has been admitted there.¹⁰

For Kempf and most modern scholars the anonymous parricide is L. Aelius Sejanus, Tiberius' Prefect of the Praetorian Guard who was executed in October 31, while for Carter the section is a non-Valerian insertion into the text, albeit one marked by ‘all the signs of bombastic lubberliness so typical of V.’¹¹ Bellemore, however, has incriminated M. Scribonius Libo Drusus the conspirator of 16, positing a correspondingly early date for Factorum et Dictorum Memorabilium. Internal evidence, which is all we can apply, is inconclusive, but Bellemore's arguments deserve scrutiny. First, Libo Drusus is alleged to have planned Tiberius' murder, whereas Tiberius' own version of events claimed that Sejanus conspired against the children of Germanicus. Second, the words ‘sidera suum vigorem obtinuerunt’ (the stars retained their strength) are taken as an overt reference to Libo Drusus' alleged consultation of astrologers. The primary reference, though, may be to Tiberius' consecrated predecessors (cf. Praef., 3. 2. 19). Third, she seeks to avoid the obvious meaning of ‘omni cum stirpe sua populi Romani viribus obtritus’ (along with all his offspring he has been crushed by the might of the Roman people) by translating ‘lock stock and barrel’.¹² However, the deaths of Sejanus and his family were celebrated on the public Fasti,¹³ whereas Libo Drusus' family, if he had one, goes unmentioned and presumably unharmed (cf. Tac. Ann. 2. 27-32). A further argument, albeit from silence, can be brought to favour a reference to Sejanus. Twice the conspirator is accused of abusing the ties of friendship (amicitiae fide extincta; violatis amicitiae foederibus). Sejanus' friendship with Tiberius had been acknowledged publicly by the Senate in 28 with the erection of an altar of friendship flanked by statues of the emperor and Sejanus (Tac. Ann. 4. 74. 2); although Libo may have owed his praetorship to Tiberius' favour and so have been bound by ties of friendship their relationship is not stressed. Indeed, were the target Libo, V. could have strengthened his attack with the charge that ties of kinship had been broken, as Libo's grandfather was probably the adoptive brother of Tiberius' mother.¹⁴

Even more precarious is use of individual examples in V. which are presumed to have a contemporary reference to events in Tiberius' reign.¹⁵ A fire in the temple of Magna Mater in ad 3 (1. 8. 11) provides the last securely datable example. V.'s description of the violence which attended the Roman theatre (2. 4. 1) is taken to reflect the troubles of 15 which required senatorial measures to quell them (cf. Tac. Ann. 1. 77. 2).¹⁶ Perhaps more probable, though, would be a reference to 23, when Tiberius banned pantomime artists from Rome (Tac. Ann. 4. 14. 3).¹⁷ V.'s criticism of Hortensius Corbio, grandson of the famous orator, who lived a life more degraded than that of a prostitute (3. 5. 4), and his acknowledgement in the preface to the next chapter of the dangerous course he has sailed in introducing such stories is held to be significant in the light of Hortensius Hortalus' appearance before the Senate in 16 to beg for a subvention from Tiberius (Tac. Ann. 2. 37-38; Suet. Tib. 47); that V.'s caution suggests a publication date close to the embarrassing incident. Briscoe, however, stresses that V.'s embarrassment is caused by his discussion of Corbio's fellatio.¹⁸

Other passages, however, can be used to suggest other, later dates and other contemporary references: from book 1 alone four examples. First, the Chief Pontiff's refusal to allow the flamen Martialis to campaign in Africa (1. 1. 2) may recall the senatorial debate of ad 20 and Tiberius' prevention of the flamen Dialis Servius Maluginensis taking up the proconsulship of Asia.¹⁹ Second, Scipio Hispalus' expulsion of the Jews from Rome in 139 (1. 3. 3) may seem particularly appropriate an example after Tiberius' expulsion of Jewish proselytizers in ad 19. Third, and closely related to the last, L. Aemilius Paullus' attack on the temple of Isis (1. 3. 4) may reflect Tiberius' destruction of her temple also in ad 19. Fourth, the survival of the statue of Servius Tullius in the temple of Fortune (1. 8. 11) may be influenced by the contemporary fate of that statue which now adorned the house of Sejanus.²⁰ Because these examples offer no definite link with the present and no first-person involvement by V., their use is at best inconclusive and any reliance placed on them, even in subsidiary arguments, is suspect. For V.'s work we must stick with the negative conclusion that its publication date remains uncertain.

3. CONTENTS, CONSTRUCTION, AND PURPOSE OF FACTORUM ET DICTORUM MEMORABILIUM

The work involves four levels of construction and organization at which the talents of V. can be scrutinized: the whole, the book, the chapter, and the section (or example). This structure provides us with the best perspective for examining the nature and purpose of the work. As it survives, V.'s work occupies 472 pages of Teubner text in nine books. It is probable that the original ending is lost, with an indeterminable amount of text: V. could be expected to have rounded off the work with a rhetorical end-piece which perhaps summarized the main themes of the work, but there is nothing.²¹ The range of topics and period of history covered are enormous, as any reading of the index shows.²²

V. has rarely been credited with organizing this collection of disparate material on any discernible plan with any degree of success, although an attempt has been made by F. Römer.²³ He emphasizes V.'s own words in the Preface as providing the basic structure: ‘Caesar, by whose heavenly forethought the virtues of which I am about to speak are fostered and vices most severely punished:’²⁴ books 3-8 constitute the virtues; book 9 the vices. He excludes from his schema books 1 and 2 because they deal with self-contained themes and have specific prefaces in which the respective themes are outlined;²⁵ from book 3 the connection of chapter to chapter is more significant than book divisions. However, the opening books, on religion and traditional customs, must be included as virtues which flourished under Tiberius, the most traditional of emperors.²⁶ In books 3 to 8 Römer believes that principles of composition can be observed, based on the system of Stoic cardinal virtues with further subdivisions such as Cicero provides.²⁷ But, for the case as he makes it, there are insurmountable problems. First, as his attempt to fit V.'s chapters to the individual virtues shows, often no pattern or principle of organization connected with the overall virtue schema can be posited. A brief survey of book 3 should illustrate this. In the preface to book 3 V. uses virtus but begins with examples of indoles, natural virtus displayed by the young, which in the introduction to 3. 2 he virtually dismisses as a prelude to the real discussion—the most forceful demonstration of which lies in fortitudo; 3. 3 patientia follows as the logical counterpart. 3. 4-5, however, although the first example (3. 4) follows naturally from the story of a slave who did not weaken under torture by the Carthaginians, and thus displayed virtus, do not fit into any philosophical scheme of virtues.²⁸ 3. 6, on famous individuals who indulged in an untraditional freedom of dress, is even harder to connect, as V.'s introduction to 3. 7 reveals: these and things similar to these are evidence of virtue taking on some degree of licence in changing what was customary. The rest of the book is readily fitted into Römer's schemes, as it consists of virtues which Cicero classes as subdivisions of fortitudo. Yet this last notion is unaccounted for by Römer, but should be crucial if V. were attempting a philosophical classification.²⁹ In the rhetorical treatise De Inventione, as many had before him and many were to do, Cicero presents a hierarchy of virtues: four main (or cardinal) virtues each with several subdivisions or categories.³⁰ V.'s classification permits no subdivisions: fortitudo is treated on the same level as constantia or patientia. Once we jettison the idea that V. adheres to the Stoic cardinal virtues which, though known to the Romans and found particularly in rhetorical theory, were too restrictive and abstract for regular Roman categories of praise,³¹ Römer's emphasis on virtus as a binding theme can survive for much of books 3-8. However, a very broad understanding of virtus and its manifestations is required, even wider than the list provided by Cicero in his description of the conflict with Catiline,³² and chapters such as 6. 2 remain problematic.³³ V.'s treatment of vices is less troublesome as it coheres in the first eleven chapters of book 9 and reaches a dramatic climax in the invective against the unnamed conspirator.³⁴ At best V. can win qualified praise for the overall composition of his work. Our discussion so far has touched upon the book unit, but more needs to be said. Of the nine books only 1 and 2 have a clear unity of theme and content; the others offer a variety of contents sometimes barely cohering. After book 3 there are no prefaces to books, so that V.'s introductions to his individual chapters, which serve as his links between different headings, give him the most prominent position to speak in his own voice and justify his subject material. If very brief connective particles are ignored,³⁵ of the 91 chapters in Kempf's division of the work,³⁶ only 36 make an indisputable link to their wider context.³⁷ This is done principally by picking up the subject of the previous chapter, for example by repeating the name of the subject or by using a relative pronoun; far less frequent is anticipation of the following chapter. It is noteworthy that two of these links cross book-divisions, as they have been preserved, with no indication of such a crossing.³⁸ This permits the suggestion that the unit of the book was less important to V. than might be thought. After book 3, in which all the chapters are linked together, only book 5 seems to be largely linked by the author. From the relative looseness of the remaining material may derive Julius Paris's notion of ten books or Gellius' miscitation,³⁹ yet to posit a redivision or a radical reorganization in order to create ‘more logical’ collocations is reminiscent of the rampant transposition of passages by nineteenth-century textual critics of, for example, Lucretius. This level of V.'s work seems the least successful.⁴⁰

V.'s treatment of examples within chapters, the third level of organization, needs less discussion here as the commentary will show V.'s various methods of transition.⁴¹ For Bloomer (Rhetoric, 26) this is the level ‘that V. has in particular organised to be understood together’. Certainly content matches carefully the subject revealed by the chapter headings. As with Suetonius' rubrics,⁴² often the reader must discern the principle of organization from the author's words: for 1. 6 V. announces ‘an account of prodigies, either favourable or adverse’, and this is what he delivers—examples 1-4 are favourable, 5-13 adverse. Beyond this, in both sequences the prodigies appear in chronological order (with exceptions caused by the narration of similar material).⁴³ Chronological considerations govern the Roman examples on 1. 2, but for 1. 5 a straightforward chronological progression (involving wholly reasonable assumptions on 1. 5. 4 and 1. 5. 7) is shattered by the final example 1. 5. 9 which is introduced as an afterthought, ‘also worthy of note’, perhaps suggested by the similarity with 1. 5. 8. In 1. 7, after two examples featuring Octavian and Julius Caesar a chronological progression is broken by 1. 7. 6.⁴⁴ Chapter 1 illustrates the problems more clearly. For this first chapter of his work, one elaborated at great length, the longest in the book and with a carefully constructed preface, some clear and impressive principle of organization might have been expected. None, however, flows from the categories of the chapter's introduction or even from the simpler ‘not only of conserving but also of increasing worship’ in 1. 1. 1: some examples of similar subject-matter are juxtaposed, e.g. those on priesthoods or magistracies abdicated because of procedural errors (1. 1. 3-6); others have an explicit chronological connection—the exploits of Albinius and Fabius Dorsuo both belonged in 390; a common reference to books may explain why M. Atilius' treachery follows Q. Petilius' destruction of the books of Numa (1. 1. 12-13). Within the first fifteen sections where worship was conserved and expanded only an intermittent gradatio (climax) can be posited: 1. 1. 3 is expressly more remarkable than 1. 1. 2; Regulus' preservation of religio surpasses all others (1. 1. 14) and climactically the Roman Senate's veneration of the gods exceeds that of the Carthaginians (1. 1. 15). The six examples of neglected worship which follow have no pattern, only the similar theme of a divinity avenging a slight to his worship or property by punishing the transgressor. However, with the nine external examples, of which the last five survive only in epitome, a pattern of antithesis emerges: the odd-numbered examples are of impiety avenged, the even of piety demonstrated; the state of survival prevents any conclusion on a possible climax. Rather than a desire for juxtaposition of the paradoxical to surprise the reader and demonstrate his virtuosity in connecting the disparate, V. more frequently demonstrates the desire for easy, obvious transitions.⁴⁵

Lastly there is the construction of the individual exempla.⁴⁶ As R. Guerrini has demonstrated,⁴⁷ the regular structure of a Valerian exemplum is introduction, substance, and conclusion. An introduction can vary from an elaborate, overtly rhetorical prelude (e.g. 1. 1. 9) to a simple particle often untranslatable into English (e.g. autem: 1. 1. 20); also common is a relative pronoun picking up something from the previous exemplum (e.g. cuius: 1. 1. ext. 4) or a demonstrative pronoun (e.g. 1. 1. 6). Failure to introduce is rare.⁴⁸ The substance speaks for itself, being the material V. has adapted from his sources to fill out his chapters.⁴⁹ Most variable is the conclusion, as sometimes none is needed, for example when similar material is presented in a series of exempla (e.g. 1. 1. 4-5, following 1. 1. 3; 1. 8 ext. 12-15). Examination of book 1 demonstrates the validity of Guerrini's structure, yet V.'s procedure and construction of the individual exempla is far from mechanical or monotonous. He exhibits, for example, great flexibility in the integration or separation of introduction and substance. Integration is usually preferred where the introduction is short⁵⁰—a key variety is where the name of the individual or state is provided in the introductory sentence or phrase and the substance begins often with a temporal⁵¹ or relative clause.⁵² Tell-tale words often mark the transition to substance, e.g. enim and nam(que),⁵³ sometimes in conjunction with the previous device. Alternatively the introduction can stand separate,⁵⁴ but connection in some form is the rule. Conclusions are the least constant element and often the hardest to differentiate from the substance.⁵⁵ They are V.'s clearest vehicle for pointing up the moral of the exemplum (e.g. 1. 1. 9, 1. 5. 1), where he can indulge in his most figured language.

For whose benefit and with what purpose did V. compile Factorum et Dictorum Memorabilium? The preface provides an initial answer to the question: ‘I have decided to make a selection from famous authors and to set them out, in order that those who wish to embrace the examples may be spared the toil of lengthy research.’ Yet the identity of these busy people remains obscure. For some scholars⁵⁶ V.'s audience is declaimers, oral performers in the recital halls of the early Principate, men who needed a wide range of material to provide the historical examples for their speeches. Others⁵⁷ prefer a wider audience which included those in public life with little leisure, whose business was politics or law. V.'s fourth-century epitomator Julius Paris justifies his abbreviation ‘since I knew that a collection of examples was necessary for lawyers no less than for declaimers,’⁵⁸ but while this may confirm a wider audience for V. in the fourth century it need say nothing about V.'s own intentions. Equally there is no internal evidence to suggest that V.'s main intended audience was not the Roman élite, but those provincials and Italians increasingly active in Rome's bureaucracy and military service.⁵⁹

As we have seen in relation to the overall structure of Factorum et Dictorum Memorabilium, V.'s words in the preface should be given due weight, so when he invokes the emperor Tiberius ‘by whose heavenly forethought the virtues of which I am about to speak are most generously encouraged and vices most severely punished,’⁶⁰ the essentially moral purpose of the work becomes clear.⁶¹Exempla were a key mode or moral education for the Romans, for whom history provided a catalogue of actions and sayings worthy of praise or blame.⁶² V. provided such a catalogue with the morals inescapably highlighted by the arrangement and by his own introductions and conclusions to the individual exempla.

A pedagogic purpose of providing models of proper declamation has also been surmised,⁶³ yet it is difficult to see most of V. in this light. Certainly V.'s introductions and conclusions, especially where he looses a host of rhetorical devices, have close parallels in the topics and sententiae of declamation, as they are recoverable from Seneca the Elder's Controversiae and Suasoriae.⁶⁴ The bulk of V., however, the material of his exempla, is different from the declamatory treatment. Seneca's Suasoriae offer excerpts from the leading declaimers of the early empire, with whose work V. would have been familiar if he lived in Rome. In a suasoria the declaimer argues a course of action for a historical (or mythological) figure, i.e. the bare historical details are wrapped up in an often extended speech, or barely feature; in a controversia speeches were made for one side or the other in a theme taken from law. Seneca only rarely preserves a continuous extract from a controversia rather than the epigrammatic highlights,⁶⁵ and the rest bears little resemblance to V. in scale. The use of historical exempla in controversiae, as Seneca presents it, is small; the references are brief, with the barest reference to the historical situation.⁶⁶ Even so a remarkable continuity of exempla between V. and Seneca is obvious: of 67 Roman exempla in Seneca, 41 appear in V.⁶⁷ V.'s wider collection would have been useful to declaimers—his introductions and conclusions would have suggested uses to which the exempla were suited; his substance would have provided appropriate names and the modicum of detail needed. While Skidmore makes a good case for a fundamental moral purpose for Factorum et Dictorum Memorabilium, that should not be divorced from the practical use to which the collection might be put. Skidmore himself argues for V.'s work being read at dinner parties of the social élite: on the basis of an anecdote in Plutarch's Moralia about a man who bored his friends with a reading from Ephorus he writes (111): ‘the popularity of works of moralistic historiography at dinners means that a work composed of morally improving episodes from history would also have been popular.’ Plutarch's Ephorus fanatic was unpopular, and there is a large, problematic jump in the argument, acknowledged by the formulation ‘would … have been popular’. Similarly an identification of the material delivered by the aretologoi, whom Suetonius records Augustus presented at dinner parties, with V.'s moralizing paragraphs has little in its favour, for in extant Latin aretologoi are characterized as longwinded, not a description easily applied to V.'s excerpted exempla, and present the virtues of the gods, not the human focus of V.

I would argue that the serious moral purpose envisaged by Skidmore should be combined with a primary audience of those involved in declamation, the advanced stage of the élite Roman's education, indeed perhaps the most practical aspect of his education. In urging upon this group the importance of morality, in providing material useful for declamation and in demonstrating through his prefaces and conclusions the way such exempla could be deployed V. is fulfilling a complex role. V. did not aim to write a mere handbook: the absence of a system of reference within the text, once the extant titles are correctly understood as post-V. (see above n. 22), and the author's own persistent entry into the text argue against this—his was a text that could be read continuously as well as mined for exempla on any specific theme.

4. VALERIUS' SOURCES

Owing to the industry largely of German scholars of the nineteenth and first half of the twentieth century much has been written on V.'s sources of information,⁶⁸ and much about his working methods is now clear, but Quellenforschung by its nature, even when the sources drawn upon are extant and full comparison is possible, permits few certainties. V. cites twenty-one authorities as if he has consulted them, ten Latin and eleven Greek.⁶⁹ Of these, twelve appear in two chapters of book 8, the Greek sources probably taken from Varro.⁷⁰ For the most part, then, as with most other ancient writers V. does not disclose his sources. In the preface he described his sources as famous authors (‘inlustribus auctoribus’), a claim which modern research supports. The theory advanced by Klotz and Bosch that V. draws principally on one or more pre-existing collections of exempla from the Ciceronian or Augustan period, the putative authors of which are Hyginus and Verrius Flaccus, is to be rejected,⁷¹ allowing V. the major role in the compilation and organization of his material. However, once V. is permitted direct consultation of his sources their number is best reduced—for example, arguments for the use of Diodorus do not survive scrutiny.⁷² What survives is use of Livy, Cicero, Varro, and Pompeius Trogus.⁷³ To this must be added the even more elusive influence of the rhetorical schools, an essentially oral and more fluid version of stories, the origins of which were probably lost when V. wrote. Perhaps, however, this is too strict an economy of sources—for example from 1. 1 there are exempla which are not found in extant Livy or Cicero: Tarquinius Superbus' punishment of M. Atilius for his treachery over the Sibylline Books (1. 1. 13); the celebration of the rites of Ceres after Cannae (1. 1. 15; contra Livy 22. 56. 4-5); and C. Terentius Varro's angering of Juno at the Circensian games, or at least the expiation of that error after Cannae (1. 1. 16). While a home for these may be posited in the many mansions of Varro's lost works we cannot exclude outright that for a period such as the Second Punic War from which many of the exempla come V. has consulted a famous historian, perhaps L. Coelius Antipater.⁷⁴ Even if he wrote with a limited range of sources V.'s use of them is not mechanical—each exemplum need not have just one source. For example, where V. takes the application of an exemplum from Cicero, he may supplement the cursory account with detail from Livy (as Bloomer demonstrates for 1. 6. 1)⁷⁵ or reject Cicero's version completely for Livy.⁷⁶ Again, while the story of Albinius seems based on Livy, the etymology of caerimonia may be Varronian.⁷⁷

Where we can compare V. with a probable source and observe how he modifies the original, his procedure is revealing. Two examples suffice:

ego vero primum habeo auctores et magistros religionum colendarum maiores nostros … qui statas sollemnisque caerimonias pontificatu, rerum bene gerundarum auctoritates augurio, fatorum veteres praedictiones Apollinis vatum libris, portentorum expiationes Etruscorum disciplina contineri putaverunt

(Cic. Har. Resp. 18)

maiores statas sollemnesque caerimonias pontificum scientia, bene gerendarum rerum auctoritates augurum observatione, Apollinis praedicationes vatum libris, portentorum depulsi‹one›s Etrusca disciplina explicari voluerunt

(V. 1. 1. 1)

V.'s debt here is inescapable; he varies Cicero by use of synonyms (praedicationes for praedictiones; depulsiones for expiationes), by slight alteration of word order (bene gerendarum rerum for rerum bene gerundarum) and exchange of adjective for noun (Etrusca for Etruscorum).

sacrificium erat statum in Quirinali colle genti Fabiae. ad id faciendum C. Fabius Dorsuo Gabino cinctu, sacra manibus gerens, cum de Capitolio descendisset, per medias hostium stationes egressus, nihil ad vocem cuiusquam terroremve motus, in Quirinalem collem pervenit ibique omnibus sollemniter peractis eadem revertens similiter constanti vultu graduque, satis sperans propitios esse deos, quorum cultum ne mortis quidem metu prohibitus deseruisset, in Capitolium ad suos rediit seu attonitis Gallis miraculo audaciae seu religione etiam motis, cuius haudquaquam neglegens gens est

(Livy 5. 46. 2-3)

eadem rei publicae tempestate C. Fabius Dorsuo memorabile exemplum servatae religionis dedit. namque Gallis Capitolium obsidentibus, ne statum Fabiae gentis sacrificium interrumperetur, Gabino ritu cinctus, manibus umerisque sacra gerens per medias hostium stationes in Quirinalem collem pervenit. ubi omnibus sollemni more peractis in Capitolium prope divina veneratione victricium armorum perinde victor rediit

(V. 1. 1. 11)

The similar techniques are evident: change of word order (statum sacrificium for sacrificium statum; Fabiae gentis for genti Fabiae), slight alteration of expressions (Gabino ritu cinctus for Gabino cinctu; sollemni more for sollemniter, ubi for ibique). At the same time the verbatim repetition of some phrases is obvious (per medias hostium stationes and in Quirinalem collem pervenit … omnibus … peractis). Yet one of the key skills of the excerptor is demonstrated in the abbreviation of the source: V. omits Livy's psychological padding to concentrate on the factual. We can imagine with some degree of plausibility that this is how V. proceeds with most of his source material and see that it is not discrepant with his claim in the preface that he could not improve upon the style of his predecessors.⁷⁸

5. TEXT AND TRANSMISSION

All Renaissance and early nineteenth-century editions of V. were superseded by C. Kempf's Editio maior of 1854, which he revised in 1888. While his work is a great improvement on what went before, his version of the stemma requires correction, as G (Bruxellensis 5336) presents a tradition independent of L and A, which were the major manuscripts, and preserves valuable readings.⁷⁹ Moreover his reports of the readings of L and A are frequently incorrect.⁸⁰ A new edition of V., which has long been a desideratum, is being prepared for Teubner by Dr J. Briscoe of Manchester University. With his kind permission a preliminary version of his apparatus criticus forms the basis for the translation of this edition.

All the extant manuscripts of book 1 have a major lacuna which extends from 1. 1 ext. 5 to 1. 4 ext. 1. For the contents of the lacuna we rely on V.'s two epitomators Julius Paris and Januarius Nepotianus.⁸¹ While both eliminate V.'s introductions their treatment of his content varies considerably. Julius Paris presents only the subject matter and remains for the most part faithful to V.'s language, although abbreviating it.⁸²

Metellus vero pontifex maximus Postumium consulem eundemque flaminem Martialem ad bellum gerendum Africam petentem, ne a sacris discederet, multa dicta urbem egredi passus non est, religionique summum imperium cessit, quo‹d› tuto se Postumius Martio certamini conmissurus non videbatur caerimoniis Martis desertis.

Metellus pontifex maximus Postumium cos. eundemque flaminem Martialem ad bellum gerendum proficiscentem urbe egredi passus non est, ne caerimoniis Martiis desertis inauspicato bellum committeretur

Nepotianus strays much further and often retains a brief moralizing conclusion, although not necessarily that drawn by V. For example 1. 1. 3 and Nepotianus' version:

Laudabile duodecim fascium religiosum obsequium, laudabilior quattuor et XX in consimili re oboedentia: a Tiberio enim Graccho ad collegium augurum litteris ex provincia missis, quibus significabat se, cum libros ad sacra populi pertinentes legeret, animadvertisse vitio tabernaculum captum comitiis consularibus, quae ipse fecisset, eaque ab auguribus ad senatum relata iussu eius C. Figulus e Gallia, Scipio Nasica e Corsica Romam redierunt et se consulatu abdicaverunt.

Tiberius Gracchus ad augurum collegium e provincia scripsit comperisse se consules non auspicato creatos. recitata in senatu epistula revocati sunt consules, Figulus ex Gallia, Scipio Nasica ex Corsica et magistratu se abdicaverunt. tanta vis religioni inerat, ut de ea crederetur absenti.

Nepotianus also changes the order of V.'s exempla on a basis that is not always obvious.⁸³ Also his promise to include some things which V. omitted is fulfilled in that part of his epitome which survives: nine exempla not in V.⁸⁴ To these may be added two exempla from the section missing from V.'s main manuscripts where Nepotianus includes two exempla absent from Julius Paris,⁸⁵ but as Julius Paris can omit exempla from V. without obvious justification,⁸⁶ certainty in these two cases is impossible. So while the lacuna has deprived us of the most original part of V.'s contribution to the missing chapters, his linkages of the individual examples, none the less, thanks to the epitomators, we can be confident of the historical exempla which V. considered illustrated the topic.

Notes

1. 2. 6. 8.

2. R. Syme (History in Ovid (Oxford, 1978), 161-2) argues for 24/25; id. ZPE [Zeitschrift für Papyrologie und Epigraphik] 53 (1983), 191, for 23/24; U. R. D. Vogel-Weidemann, Die Statthälter von Africa und Asia in den Jahre 14-68 n. Chr. (Bonn, 1982), 259, 495 suggests 25/26. The identification is questioned by C. J. Carter, ‘Valerius Maximus’ in T. A. Dorey (ed.), Empire and Aftermath: Silver Latin II (London, 1975), 31-2; cf. J. Bellemore, ‘When did Valerius Maximus write the Dicta et facta memorabilia?’ Antichthon, 23 (1989), 76-7.

3. 4. 11. 4. RE [Real-Encyclopädie der klassischen] viiia, 90; contra Carter, ‘Valerius Maximus’, 52 n. 26.

4. Praef.: ‘my littleness’ (mea parvitas).

5. C. J. Skidmore, Practical Ethics for Roman Gentlemen: The Work of Valerius Maximus (Exeter, 1996), 113-17, argues from the first person references to ancestral images in 5. 5 praef. that V. was related to the patrician Valerii Messallae, a gens which revived several old cognomina in the early years of the empire.

6. Praef.; 2. praef.; 5. 5. 3, 8. 13, praef.; 9. 11 ext 4. The penultimate speaks of the emperor's old age, ‘ad longissimos humanae condicionis terminos’.

7. The Pompeius whom Seneca records was starved to death by Caligula (Tranq. 11. 10) was not the consul of 14, but probably his son (R. Syme, The Augustan Aristocracy (Oxford, 1986), 132 n. 30, 414 n. 72), hence there is no need to advance the publication of book 4 and following to the beginning of Claudius' reign (contra [Pierre] Constant [in his edition of Valerius (Paris, 1935)], p. ii) in order to have 4. 7 ext. 2, where V. probably laments the death of his friend Pompeius, refer to this Pompeius (cf. [R.] Combès [Valère Maxime: Faits et dits mémorables livres I-III], 9).

8. Despite the trenchant criticism of Carter, ‘Valerius Maximus’ 31-3, the basic schedule laid down by Kempf is still acceptable (cf. W. M. Bloomer, Valerius Maximus and the Rhetoric of the New Nobility (London, 1992), 1).

9. ‘tu enim prisca religione consecratos Vestae focos incolis, tu Capitolinae Iunonis pulvinaribus incubas, tu Palatii columen, augustos penates sanctissimumque Iuliae genialem torum adsidua statione celebras.’

10. 9. 11 ext. 4.

11. 33. Cf. R. Syme, ZPE 53 (1983), 192-3: ‘a patent insertion.’ However, the notion of a second edition has nothing to recommend it, as it rests solely on the perception that the Sejanus passage is awkwardly inserted into its present context. But the passage is a well-rounded conclusion to the whole chapter, so J. Briscoe, ‘Some Notes on Valerius Maximus’, Sileno, 19 (1993), 401.

12. 79 n. 49. F. G. B. Millar, JRS [Journal of Roman Studies] 83 (1993), 4 inclines to Bellemore's identification: ‘may well.’

13. Fasti Ostienses; see L. Vidman, Fasti Ostienses (Prague, 1982), 64-5.

14. Cf. Tac. Ann. 2. 27. 2: ‘consobrinos Caesares.’ For Libo's familial connections, see E. J. Weinrib, HSCP [Harvard Studies in Classical Philology] 72 (1967), 247-78; Syme, The Augustan Aristocracy, stemmata xiv-xv.

15. Attempted by Bellemore, ‘When did Valerius Write the Dicta?’, 75-6.

16. See W. J. Slater, CA [Classical Antiquity] 13 (1994), 125 n. 32.

17. Cf. E. J. Jory, BICS [Bulletin of the Institute of Classical Studies of the University of London] 28 (1981), 152.

18. ‘Some Notes’, 402.

19. C. Santini, ‘Echi di politica religiosa tiberiana in Valerio Massimo’, GIF [Giornale italioano di filologia] 39 (1987), 192-4, discusses the debate as an example of the religious attitudes of Tiberius' reign—V.'s example was not ‘casuale, ma è perfettamente calzante ed ha tutto il sapore della conferma di una scelta di politica religiosa di Tiberio in una questione peraltro controversa.’

20. Santini ‘Echi’, 194, concentrates on another aspect of the same example: the statue of Quinta Claudia may have been suggested by the miraculous preservation of a statue of Tiberius during a fire on the Caelian hill which Tacitus records under 27 (Ann. 4. 64. 3).

21. Cf. Carter, ‘Valerius Maximus’, 29. V.'s earliest epitomator Julius Paris writes of ten books, but this is most likely an error (e.g. Helm, RE viiia, 115). Certainly we do not suspect a large missing book. Paris's error has been attributed to the existence of an introductory volume which listed the contents of the whole work (cf. Helm, RE viiia, 114), but, if any credence is to be attached to Paris's words, we should posit a different arrangement of V.'s material. Aulus Gellius, for instance, gives a reference from our book 8 (8. 1 amb. 2) as from book 9, but Gellius' citations are far from infallible (see D. Wardle, AClass [Acta classica: proceedings of the Classical Association of South Africa] 37 (1993), 93-4).

22. The headings as they stand within the text today are the work of early copyists (see W. Thormeyer, De Valerio Maximo et Cicerone quaestiones criticae (Diss. Göttingen, 1902), 33-5; Helm, RE viiia, 97), as shown by the fact that the regular formula ‘about x’ stands outside the syntax of V.'s Latin and frequently interrupts his connections between chapters and books.

23. e.g. Helm, 95 RE viiia, who grants that he made an effort, but that any principles of organization are often in conflict—6. 5 seems oddly separated from 8. 1-6, or 3. 6 from 9. 1. Carter's verdict (‘Valerius Maximus’ 27-8) is most negative: ‘There is no obvious or subtle unity to these books, individually speaking, and their collective content has no aesthetic pattern or logical coherence. Closely related topics or similar stories can be widely separated between books. This is again characteristic of the whole work and Books VI to VIII are even less homogeneous than the previous five … Book VII is a potpourri … VIII a mixed bag etc.’ But F. Römer, ‘Zum Aufbau der Exemplasammlung des Valerius Maximus’, WS [Wiener Studien] 103 (1990), 99-107. I have not seen the thesis he refers to: G. Schmied, Die literarische Gestaltung der Factorum et Dictorum memorabilium Libri novem des Valerius Maximus (Vienna, 1990).

24. ‘Caesar, invoco, cuius caelesti providentia virtutes, de quibus dicturus sum, benignissime foventur, vitia severissime vindicantur.’

25. Römer, ‘Zum Aufbau’, 100. The preface to book 1 proper, which V. adapts from Cicero, does not really cover the last two chapters, the contents of which lie outside the state religion.

26. Skidmore, Practical Ethics, 64-71, rightly stresses the basic link between religion and the morality V. wishes to encourage.

27. De Inventione 2. 159-64. Prudentia is demonstrated in memoria, intelligentia, and providentia; iustitia in religio, pietas, gratia, observantia, vindicatio, and veritas; fortitudo in magnificentia, fidentia, patientia, and perseverantia; and temperantia in continentia, clementia, and modestia.

28. Combès (29-30) attempts to relate these two contrasting chapters to the contents of Rhet. Her. 3. 13.

29. Combès (24-46) provides a detailed account of the organization and content of the nine books, but his main idea, that they are governed by four cardinal virtues, does not withstand scrutiny.

30. See above, n. 7. H. F. North, in The Classical Tradition: Literary and Historical Studies in Honor of Henry Caplan (Ithaca, NY, 1966), 165-83 offers a concise history of these classifications. Cf. A. F. Wallace-Hadrill, Historia 30 (1981), esp. 300-4. The flexibility of the canons, especially the frequency with which εὐσέβεια (religio) is included or not may confirm the view that book 1 clearly belongs among the virtues.

31. Cf. C. J. Classen, Gymnasium 95 (1988), 289-302.

32. Cat. 2. 25.

33. Cf. Bloomer, Rhetoric, 54-5. The conclusion to 6. 2 ‘but because I have the intention of running through stages of human life’ (sed quia humanae vitae partes persequi propositum est) appears to indicate a programme which was not carried out. In book 6 the links between chapters in the order they have been preserved are non-existent or weak Chapters 2 and 4 could be removed to create a more homogeneous whole.

34. See p. 3 [of Introduction in Valerius Maximus: Memorable Deeds and Sayings, Book I.]

35. e.g. 2. 3: etiam; 6. 2 praef.: autem; 6. 5 praef.: quoque.

36. By contrast with Kempf, I do not treat 7. 8. 5f. as a separate category—indeed V.'s introduction to 7. 8 is broad enough to include examples 5 to 9. 9. 13 is more problematic; after the first unexceptional foreign examples of cupiditas vitae V. provides what appears to be another preface, ‘I shall now mention those who …’, to three examples. These do demonstrate cupiditas vitae, but their introduction marks them out. Cf. 8. 1. amb. 1 which introduces a class of examples not mentioned in the preface to 8. 1. Combès (46) has 104 chapters, without indicating his divergence from Kempf.

37. I count as definite the following (* denotes a forward announcement): 1. 7. 1, 1. 8 praef., 2. 4. 1, 2. 8 praef., 2. 9 praef., 2. 10 praef., 3. 2 praef., 3. 3 praef., 3. 5 praef., 3. 6 praef., 3. 7 praef., 3. 8 praef., 4. 2 praef., 4. 5 praef., 4. 6 praef., 5. 1 praef., 5. 4 praef., 5. 5 praef.*, 5. 6 praef., 5. 8 praef., 5. 9 praef., 5. 10 praef., 6. 6 praef., 7. 1 praef., 7. 2 praef., 7. 3 praef., 7. 7 praef.*, 7. 8 praef., 8. 2 praef., 8. 4 praef., 8. 8 praef., 8. 10 praef., 8. 13 praef., 9. 2 praef., 9. 9 praef., 9. 11 praef., 9. 13 praef. This list could be expanded if 1. 2-4 existed in more than epitome. Between 3. 3 and 3. 4 also a link may exist: the conclusion of 3. 3 ext. 7 contains a phrase ‘humili loco nati’ which appears in the introduction to the subject matter of 3. 4 and 3. 5. Probably Kempf's paragraphing should be amended and ‘quo evenit …’ be called 3. 4. praef. I have excluded 5. 7. praef. as a link to 5. 6, for, although the appearance of pii may suggest a connection with the manifestations of pietas discussed in 5. 4-6, it refers rather to indulgentia, the subject matter of the next chapter.

38. 5. 1. praef., 7. 1 praef.

39. See above, n. 21.

40. While Bloomer's view (Rhetoric, 25) that ‘when the overt structures that purport to tie a work together are the author's only entry into the text and various rhetorical junctures, often metonymic, the structure of the whole will not seem organised,’ V.'s failure to connect over 50 per cent of his chapters serves to damn him.

41. And see Bloomer, Rhetoric, 28-54.

42. See G. B. Townend, in T. A. Dorey (ed.), Latin Biography (London 1967), 85 f.

43. See Bloomer, Rhetoric, 29-30.

44. V. may have deliberately sacrificed chronology to create an easy link with 1. 7. 7.

45. Contra Bloomer, Rhetoric, 25 f.

46. How many of these V. distinguishes is disputed: Bloomer, Rhetoric 1, has 967—a simple sum of the numbered exempla in Kempf's edition. This understates the total considerably. I identify 1,032 from closer examination of Kempf's paragraphing and subdivision of his numbered exempla. Even this total could be increased if cases such as 6. 1. 13 were considered as five individual examples (cf. 9. 2. 1). Combès (22 n. 2) identifies 989, 663 Romana, 326 externa.

47. Studi su Valerio Massimo (Pisa, 1981), 13-28.

48. e.g. 1. 1. 16 ext. 1, 1. 8. 9.

49. Combès (51-3) has useful remarks on the variety of style V. can use for his subject-matter.

50. Excluding exempla where the introduction is a simple connective i.e.: 1. 1. 2, 1. 1. 5, 1. 1. 7, 1. 1. 13, 1. 5. 4, 1. 5. 6, 1. 6. 3, 1. 6 ext. 2, 1. 7 ext. 6, 1. 7 ext. 8, 1. 8. 4, 1. 8. 5, 1. 8. 12(b), 1. 8 ext. 13, 1. 8 ext. 16.

51. 1. 1. 1(b), 1. 1. 17, 1. 1. 21, 1. 5. 2, 1. 5. 3, 1. 5. 9, 1. 6. 3, 1. 7 ext. 3, 1. 8. 1(a), 1. 8 ext. 6, 1. 8 ext. 8.

52. 1. 1. 1(c), 1. 1. 9, 1. 1. 14, 1. 1. 15, 1. 1. 18, 1. 1. 19, 1. 1 ext. 2, 1. 5. 1, 1. 5. 8, 1. 6. 2, 1. 6. 5(a), 1. 6. 7, 1. 7. 8, 1. 7 ext. 5, 1. 7 ext. 7, 1. 7 ext. 9, 1. 8. 8, 1. 8. 11(?), 1. 8. 12, 1. 8 ext. 11.

53. 1. 1. 8, 1. 1. 10, 1. 1. 11, 1. 1. 17, 1. 1. 21, 1. 1 ext. 1, 1. 1 ext. 4, 1. 5. 5, 1. 5. 9, 1. 6. 8, 1. 6. 10, 1. 6. 13, 1. 6 ext. 1(b), 1. 7. 2, 1. 7. 3, 1. 7. 6, 1. 7 ext. 1-4, 1. 7 ext. 9, 1. 8 ext. 3, 1. 8 ext. 6-7, 1. 8 ext. 12, 1. 8 ext. 18.

54. 1. 1. 1, 1. 1. 3, 1. 1 ext. 3, 1. 5 ext. 1, 1. 6. 1, 1. 6. 4, 1. 6. 5(b), 1. 6. 11, 1. 7. 4, 1. 7. 7, 1. 7 ext. 10, 1. 8. 1, 1. 8. 2?, 1. 8. 3, 1. 8. 6, 1. 8. 7, 1. 8 ext. 2, 1. 8 ext. 10, 1. 8 ext. 18.

55. e.g. 1. 1. 18, 19 where result clauses are at one level a continuation of the narrative, but in another are the conclusion.

56. e.g. K. Alewell, Über das rhetorische Paradeigma: Theorie, Beispielsammlungen, Verwendung in der römischen Literatur der Kaiserzeit (Diss. Kiel, 1913), 38.

57. e.g. Skidmore, Practical Ethics, 159 f.

58. Praef.: ‘exemplorum conquisitionem cum scirem esse non minus disputantibus quam declamantibus necessariam.’

59. The suggestion of Bloomer, Rhetoric, 12-13. Skidmore (Practical Ethics, 107) dismisses Bloomer's idea as ‘too vague to be convincing’. While there is no internal evidence to prove or indicate clearly such an intended audience, the readiness with which municipia and coloniae within Italy and the empire accepted and utilized the imperially promulgated language of virtues, and the importance placed on widespread publication of imperial documents (as seen in the tabula Siarensis and the s. c. de Pisone patre), suggests that V.'s work would have found many readers, even if we follow Skidmore (Practical Ethics, 103) and restrict V.'s audience to those of Equestrian status and above.

60. For the Latin see above, n. 24.

61. Cf. Skidmore, Practical Ethics, esp. 53-82 and the earlier views listed by M. Fleck, Untersuchungen zu den Exempla des Valerius Maximus (Diss. Marburg, 1974), 1-2 n. 1.

62. See e.g. T. P. Wiseman, Clio's Cosmetics (Leicester, 1979), 37-40.

63. Bloomer, 8-9: ‘like Seneca [the Elder] V. offers not theory but a piece of declamation itself … it is the product of that most occasional oral genre, declamation, worked out on a grand written scale.

64. See B. W. Sinclair, Prometheus, 10 (1984), 141-6.

65. Contr. 2. 7. 1-9.

66. e.g. Contr. 1. 1. 5, 1. 5. 3, 1. 8. 12.

67. e.g. Contr. 1. 5. 3: Verginia (cf. 6. 1. 2), Lucretia (cf. 6. 1. 1), the Sabine women (cf. 2. 4. 4); 2. 1. 8: Aelius Tubero (cf. 4. 3. 7, 4. 4. 9), C. Fabricius Luscinus (cf. 4. 3. 6); 2. 2. 4: Sulla (cf. 9. 2. 1); 3. 9: Servius Tullius (cf. 3. 4. 3); 7. 2. 6: C. Marius (cf. 2. 10. 6).

68. Kempf's editio maior of 1854 laid the essential foundations for Valerian Quellenforschung by collecting with remarkable accuracy and completeness the parallels to V.'s exempla. See also C. Elschner, Quaestiones Valerianae (Diss. Berlin, 1864); F. Zschech, De Cicerone et Livio Valeri Maximi fontibus (Diss. Berlin, 1865); H. E. Dirksen, Die historische Beispielsammlung des Valerius Maximus und die beiden Auszüge derselben (Leipzig, 1871); M. Kranz, Beiträge zur Quellenkritik des Valerius Maximus (Diss. Posen, 1876); B. Krieger, Quibus fontibus Valerius Maximus usus sit in eis exemplis enarrandis, quae ad priora rerum Romanarum tempora pertinent (Diss. Berlin, 1888); S. Maire, De Diodoro Siculo Valeri Maximi auctore (Diss. Schöneberg, 1898/9); W. Thormeyer, De Valerio Maximo et Cicerone quaestiones criticae (Diss. Göttingen, 1902); A. Klotz, Hermes 44 (1909), 198-214; C. Bosch, Die Quellen des Valerius Maximus: ein Beitrag zur Erforschung der Literatur der historischen Exempla (Stuttgart, 1929); A. Ramelli, Athenaeum 14 (1936), 117-52; R. Helm, ‘Valerius, Seneca und die Exemplasammlung’, Hermes, 74 (1939), 130-54; R. Helm, ‘Beiträge zur Quellenforschung bei Valerius Maximus’, RhM [Rheinisches Museum] 89 (1940), 241-73; A. Klotz, Studien zu Valerius Maximus und den Exempla (Munich, 1942); F. R. Bliss, ‘Valerius Maximus and his Sources: A Stylistic Approach to the Problem’ (Diss. NC., 1951); M. Fleck, Untersuchungen zu den Exempla des Valerius Maximus (Diss. Marburg, 1974); G. Maslakov, ANRW [Aufstieg und Niedergang der römischen Welt] ii. 32. 1 (Berlin, 1984), 437-96; and Bloomer, Rhetoric, 59-146.

69. See list at Bloomer, Rhetoric, 63; cf. Kempf, 625, for a fuller list justifiable in different terms.

70. Fleck, Untersuchungen, 8.

71. The articles of Helm (and Bloomer, Rhetoric, 64-7), and Skidmore (Practical Ethics, 47-8) expose the weaknesses of the theory. On the range of earlier compilers of exempla, see Skidmore, ibid. 45-7, and Combès, 20-1.

72. See Bloomer, Rhetoric, 79-107 for the demolition of Maire's thesis.

73. For Sallust as a limited source, see Bloomer, Rhetoric, 108-13.

74. At 1. 7. 6, however, where V. appears to cite Coelius, he is merely elaborating on what Cicero writes in De Divinatione 1. 56.

75. Bloomer, Rhetoric, 36.

76. e.g. 1. 6. 3 (cf. Bloomer, Rhetoric, 37).

77. 1. 1. 10; see ad loc.

78. See Bliss, Valerius Maximus and his Sources, for a lengthy demonstration and Bloomer, Rhetoric, 232-43.

79. The importance of G was first recognized by D. M. Schullian (CP [Classical Philology] 32 (1937), 349-59). A detailed reassessment has been made by C. J. Carter, ‘The Manuscript Tradition of Valerius Maximus’ (Diss. Cambridge, 1968). The new Budé text by Combès marks a step backwards in Valerian studies, as he ignores, without comment, the importance and existence of G; his extreme conservatism also leads to the retention of impossible readings from L and A. Moreover his changing of sigla that have been standard since Kempf (A becomes B and B becomes V) is unnecessarily confusing.

80. So Carter; cf. L. R. Reynolds, Texts and Transmission (Oxford, 1983), 423.

81. What little is known of these is set out in Handbuch der lateinischen Literatur der Antike: Restauration und Erneuerung, ed. R. Herzog (Munich, 1989), 193-5.

82. e.g. 1. 1. 2 and its parallel.

83. 1. 1. 13 appears after 1. 1. 9, but before 1. 1. 1 and 1. 1. 3 etc; see Kempf, 593-6.

84. Praef.: ‘nonnulla praetermissa conectam’; 8. 24, 34; 15. 1, 7; 17-20; 21. 3.

85. In my numeration 1. 2. 3a and 1. 4. praef.

86. e.g. from book 1: 6. 5; 7. 2, 7 ext. 5; 8. 7, 8 ext. 13-15, 18.